Modelling and Optimization of Sorption-Enhanced Biomass Chemical Looping Gasification Coupling with Hydrogen Generation System Based on Neural Network and Genetic Algorithm

Xudong Wang,Sheng Wang,Baosheng Jin,Zhong Ma,Xiang Ling
DOI: https://doi.org/10.1016/j.cej.2023.145303
IF: 15.1
2023-01-01
Chemical Engineering Journal
Abstract:Chemical looping gasification of biomass (BCLG) can realize the production of pure syngas without an extra purification process. By coupling steam oxidation of oxygen carrier in BCLG, pure hydrogen can be generated meanwhile, which is a promising clean energy. An enhanced BCLG process coupling with hydrogen generation is constructed in this work, aiming to realize effective co-production of syngas and hydrogen. The coupled effects of temperature and material flows of gasifier on the gas yield, lower heating value (LHV) of syngas and gasification efficiency are numerically investigated. The effects of temperature and steam flowrate in oxidizer are changed to investigate their effects on hydrogen production. Based on the simulation results, an accurate back-propagation neural network (BPNN) is trained and tested for the performance prediction of this syngas and hydrogen coproduction system. The prediction accuracy of this BPNN model is quite high with a correlation coefficient of 0.99967. Finally, the optimization of this system is conducted based on the BPNN model and genetic algorithm (GA) to make the H2/CO ratio close to 2 and maximize hydrogen production simultaneously.
What problem does this paper attempt to address?